AI Paper Atlas

Research map / 3D vision / Human pose estimation

Multi-level Motion Attention for Human Motion Prediction

Wei Mao, Miaomiao Liu, Mathieu Salzmann, Hongdong Li

IJCV 2021 · 3D vision › Human pose estimation · approach: Human motion prediction · 99 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add post-fusion of joint, part and pose attention predictors (ablation)Motion attention model · H3.6M long-term average 1000msMPJPE (mm): 112.1 -> 110.1Improved
add post-fusion (ablation)Motion attention model · H3.6M short-term average 400msMPJPE (mm): 58.3 -> 57.3Improved
add post-fusion (ablation)Motion attention model · H3.6M Walking 1000msMPJPE (mm): 58.1 -> 57.1Improved
add post-fusion (ablation)Motion attention model · AMASS-BMLrub 1000msMPJPE (mm): 67.2 -> 65.8Improved
add post-fusion (ablation)Motion attention model · 3DPW 1000msMPJPE (mm): 73.7 -> 71.8Improved
add motion attention to the LTD predictor (10-frame vs 50-frame observation baseline)GCN predictor · H3.6M Walking 1000msMPJPE (mm): 70.7 -> 57.1Improved
add motion attention to the GCN predictorLTD-10-10 · H3.6M long-term average 1000msMPJPE (mm): 114.0 -> 110.1Improved
use 100 observed frames at test time instead of 50 (ablation)Motion attention model (trained with 50 frames) · H3.6M selected sequences with long-range repetition, 1000msMPJPE (mm): 121.1 -> 116.9Improved
use 100 observed frames at test time instead of 50 (ablation)Motion attention model · H3.6M selected sequences, 80msMPJPE (mm): 10.7 -> 10.7No clear effect
attention only vs LPJ attentionPose motion attention · H3.6M 4 actions average 400msMPJPE (mm): 41.2 -> 37.9Improved
replace LTD-10-10 by fused attention modelMotion Att. + Post-fusion · AMASS-BMLrub 1000msMPJPE (mm): 91.2 -> 65.8Improved
fusion vs LTD-10-10 at very short termMotion Att. + Post-fusion · AMASS-BMLrub 80msMPJPE (mm): 10.3 -> 11.0Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy